The research will investigate the distribution and characteristics of the test statistics associated with the Shine-Bower model, the Gentile model and the Hartmann modification of the Gentile model for n equals 1 designs. Estimates of alpha and power will be given for each test and robustness to violation of assumptions will be noted. Variables to be studied are patterns of correlations, various learning curves, various shapes of population distributions, and equal/unequal variances. Monte-Carlo methods will be used to examine the performance of the test statistics using situations patterned after actual data from real experiments.